Hehl et al. BMC Genomics (2015) 16:66 DOI 10.1186/s12864-015-1225-x

RESEARCH ARTICLE Open Access

Asexual expansion of Toxoplasma gondiimerozoites is distinct from tachyzoites and entailsexpression of non-overlapping gene families toattach, invade, and replicate within felineenterocytesAdrian B Hehl1*, Walter U Basso1†, Christoph Lippuner1,2†, Chandra Ramakrishnan1, Michal Okoniewski3,Robert A Walker1,4, Michael E Grigg5, Nicholas C Smith4 and Peter Deplazes1

Abstract

Background: The apicomplexan parasite Toxoplasma gondii is cosmopolitan in nature, largely as a result of its highlyflexible life cycle. Felids are its only definitive hosts and a wide range of mammals and birds serve as intermediatehosts. The latent bradyzoite stage is orally infectious in all warm-blooded vertebrates and establishes chronic,transmissible infections. When bradyzoites are ingested by felids, they transform into merozoites in enterocytes andexpand asexually as part of their coccidian life cycle. In all other intermediate hosts, however, bradyzoites differentiateexclusively to tachyzoites, and disseminate extraintestinally to many cell types. Both merozoites and tachyzoitesundergo rapid asexual population expansion, yet possess different effector fates with respect to the cells and tissuesthey develop in and the subsequent stages they differentiate into.

Results: To determine whether merozoites utilize distinct suites of genes to attach, invade, and replicate within felineenterocytes, we performed comparative transcriptional profiling on purified tachyzoites and merozoites. We usedhigh-throughput RNA-Seq to compare the merozoite and tachyzoite transcriptomes. 8323 genes were annotated withsequence reads across the two asexually replicating stages of the parasite life cycle. Metabolism was similar betweenthe two replicating stages. However, significant stage-specific expression differences were measured, with 312transcripts exclusive to merozoites versus 453 exclusive to tachyzoites. Genes coding for 177 predicted secretedproteins and 64 membrane- associated proteins were annotated as merozoite-specific. The vast majority of knowndense-granule (GRA), microneme (MIC), and rhoptry (ROP) genes were not expressed in merozoites. In contrast, a largeset of surface proteins (SRS) was expressed exclusively in merozoites.

Conclusions: The distinct expression profiles of merozoites and tachyzoites reveal significant additional complexitywithin the T. gondii life cycle, demonstrating that merozoites are distinct asexual dividing stages which are uniquelyadapted to their niche and biological purpose.

Keywords: Toxoplasma gondii, Apicomplexa, Coccidia, Cat, Enteroepithelial development, Merozoite, Schizont,Comparative transcriptomics, Surface antigen, Stage-specific gene expression

* Correspondence: adrian.hehl@uzh.ch†Equal contributors1Institute of Parasitology-University of Zurich, Winterthurerstrasse 266a, Zürich8057, SwitzerlandFull list of author information is available at the end of the article

© 2015 Hehl et al.; licensee BioMed Central. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly credited. The Creative Commons Public DomainDedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article,unless otherwise stated.

Hehl et al. BMC Genomics (2015) 16:66 Page 2 of 16

BackgroundToxoplasma gondii is an intracellular zoonotic parasitethat chronically infects 30% of the world’s human popu-lation [1]. It has a complex life cycle, infecting a widerange of mammals and birds as intermediate hosts butwith felids as the only definitive hosts. Intermediatehosts can become infected through ingestion of oocystsshed into the environment via cat feces or by ingestingtissue cysts in meat or viscera [2]. During acute infectionof naïve intermediate hosts, sporozoites from oocysts orbradyzoites from tissue cysts differentiate quickly intotachyzoites, which divide rapidly and disseminate through-out the host’s body. Tachyzoites are, however, controlledefficiently by cell mediated immunity and the parasite re-verts to the slowly replicating or quiescent bradyzoiteform, which resides within tissue cysts that are particularlyabundant in brain and heart muscle but are also presentthroughout skeletal muscle [3]. When cats ingest tissuecysts, bradyzoites can take a different developmental path-way. They are released from their cysts and invade entero-cytes of the small intestine, transforming into schizonts[4,5]. The parasite population that develops in cat entero-cytes undergoes a classical coccidian cycle involving sev-eral rounds of asexual division and amplification followedby differentiation into macro- and microgamonts, thedimorphic stages of sexual development. Microgametesfertilize macrogametes, producing diploid zygotes thatsubsequently develop into unsporulated oocysts that areexcreted in the feces of the cat. The sexual phase con-tinues within the oocyst as meiosis ensues, followed bymitosis to produce infectious sporozoites, encased withinsporocysts inside the oocysts.The readily culturable, rapidly dividing tachyzoite is

the best studied form of T. gondii by far – there is abun-dant information about cell cycle, metabolism and hostparasite interactions for this stage [6]. In contrast, themerozoite, which is the other rapidly dividing asexualform of T. gondii that ultimately generates hundreds ofmillions of gametes, is the least well studied develop-mental stage. This is largely because merozoites are notcultivatable in vitro and difficult to access in vivo. Thus,investigation of the molecular mechanisms governinginitiation of parasite amplification preceding the deve-lopment of sexual stages in cats has been severely ham-pered. The need for a more detailed understanding ofthe development in the definitive host is underscored bythe fact that infected cats shed hundreds of millions ofoocysts that can remain infectious for over a year to awide range of highly susceptible intermediate hosts,including humans [7]. Here, we developed improvedprotocols for cat infection, parasite isolation, and nextgeneration sequencing to close this knowledge gap bybuilding a transcription profile for the merozoite stageof enteroepithelial development. Using genome-wide

comparative transcriptomics, we show that merozoitesexpress distinct gene families in a stage-specific fashion,and fail to express the majority of annotated ROP, GRAand MIC proteins which are upregulated during tachy-zoite replication. Among the most highly differentiallyregulated parasite proteins were several large gene fa-milies, including those coding for SRS proteins found onthe parasite cell surface. Other key genes expressed bytachyzoites whose products are known to facilitate mo-tility, host seeking, attachment, invasion, and remodelingof the parasitophorous vacuole (PV) within the para-sitized host cell were not expressed in merozoites. Thisstrongly suggests that merozoites are biologically distinctand utilize a different suite of genes that are necessaryfor asexual expansion within feline enterocytes prior togamont development.

Results and discussionPurification of merozoites from feline enterocytes forRNA-Seq analysisT. gondii parasite preparations were generated fromenterocytes from an infected cat at onset of patency (5dpost infection). The enterocyte cell layer containing rep-licating parasites as shown by IFA (Figure 1A) from tworegions of the rinsed and opened small intestine were se-lectively harvested by mechanical stripping (Figure 1B).Repeated centrifugation and re-suspension of this mater-ial with ice-cold 0.05% Tween80 in PBS yielded micro-scopically pure fractions of extracellular merozoites (twobiological replicates). The released merozoites showeddifferential staining with Diff-Quick (Figure 1C); nogametocyte stages were observed in this fraction by mi-croscopy. However, even though there is no evidence forthis in the RNA-Seq data (see also in Methods) we can-not completely exclude minor contributions from sexualstages (macrogametes, microgametes) to the RNA poolselected for analysis. Nevertheless, based on the over-whelming majority of merozoites in the sample at thisearly time point of infection we will henceforth refer tothis fraction as “merozoites”. Although further Percollgradient purification yielded merozoites that were com-pletely free of host material, we used the detergent-washed preparations containing minor contaminationwith host RNA (Figure 1D) for sequencing to avoid anychanges to the parasite transcriptome due to additionalmanipulations.

Merozoites possess highly significant stage-specific mRNAexpression differences from tachyzoitesRead mapping revealed a significant number of uniquereads for 8323 genes after removal of reads mapping totRNA and ribosomal RNA sequences, and normalization.Calculation of RPKM values (i.e. mapped reads per kilobase per million reads in dataset) to normalize for dataset

Figure 1 Experimental infection, parasite isolation, and RNA preparation. A) Immunofluorescence assay using sheep immune serum againstT. gondii revealed numerous infected enterocytes. Nuclei are counterstained with DAPI. Shown in the top panel is a 20x magnification of asection of the small intestine (bar = 100 μM) where villi are visible. The bottom panel at 100x magnification shows a schizont containing severalmerozoites (scale bar = 5 μm). B) Enterocytes containing CZ-strain merozoite stages were stripped away selectively, leaving the villus structureand the cells of the lamina propria intact. Histology section (Hematoxilin & Eosin stained) showing stripped villi at day 5 post infection. C) Microscopicexamination of parasites in the detergent washed preparation showed only merozoite stages. D) Quality control of total RNA extracted from parasitepreparations separated on an Agilent RNA 6000 Pico Chip. The bands generated by host 28S/18S ribosomal RNA (arrowheads) and parasite 26S/18Sribosomal RNA (arrows) as well as a size marker are indicated. The samples analyzed were: raw, unprocessed material from scraped intestinal lining;Tween 80, material that was syringe-passaged and washed twice with PBS/0.05% Tween 80; Percoll, highly enriched parasite fraction after detergenttreatment and Percoll gradient centrifugation; Tachy, RNA prepared from tachyzoites grown in cell culture with human foreskin fibroblasts as host cells.

Hehl et al. BMC Genomics (2015) 16:66 Page 3 of 16

size and gene length revealed 7148 genes with thresholdRPKMs ≥10 in at least one of the two life-cycle stages.Comparative analysis (tachyzoite versus merozoite mappedreads per gene) was performed using DESeq [8] to detectdifferential gene expression. We used a high thresholdof ≥ 8-fold difference in mRNA levels (measured as nor-malized averaged mapped sequence reads per gene fromtwo datasets each for tachyzoites and merozoites) toidentify stage-specifically regulated genes [9]. The rigorousanalytical approach revealed significant stage-specific ex-pression differences for approximately 10% of genesannotated in ToxoDB (Figure 2A). Applying the ≥ 8-folddifference threshold we identified 453 genes with tachyzoite-specific, versus 312 genes with merozoite-specific ex-pression. This strategy therefore revealed only the moststrongly stage-specifically regulated genes; maximal dif-ferences of RNA expression values of 54-fold (ORF

TGME49_295662, merozoite-specific) and 404-fold (ORFTGME49_215980, tachyzoite-specific) were measured.RNAs with ≥ 8-fold difference in abundance at each

life cycle stage were parsed, and distributions into elevengene product or functional categories are shown as piecharts (Figure 2B, Additional file 1: Table S1A, B). Theglobal comparison identified hypothetical genes as themost abundant set of stage-specific genes (146/47% inmerozoites versus 294/64% in tachyzoites) expressedbetween the two developmental stages examined. TheRNA-Seq data will also serve to annotate, and therebysignificantly reduce, the number of hypothetical genesfound in both samples. Genes related to metabolism orthe cytoskeleton were not differentially regulated, withonly a few exceptions. In contrast, the majority of genesimplicated in host parasite interactions, such as secretedand surface proteins, were regulated in a stage-specific

Figure 2 Global comparative transcriptome analysis. A) Scatterplot depicting expression levels as mapped read (DESeq values) of 8323identified genes in CZ tachyzoites (Tz) and merozoites (Mz). The threshold for stage-regulated expression was set at ≥8-fold difference. B) Piecharts show parsing of 453 tachyzoite and 312 merozoite stage-regulated genes into functional categories.

Hehl et al. BMC Genomics (2015) 16:66 Page 4 of 16

fashion (Figure 3). For example, significantly more sur-face (SRS) proteins were expressed in merozoites than intachyzoites (40 versus 12), and their expression was re-stricted exclusively to the merozoite developmentalstage. In striking contrast, the majority of microneme-,rhoptry-, and dense granule-specific secreted proteinswere not expressed in merozoites, suggesting the exist-ence of an unannotated suite of MIC, ROP and GRAproteins that promote merozoite asexual replication.

Merozoite specific gene familiesSeveral large, unannotated gene families containingmany predicted secreted and/or membrane-associatedproteins, referred to as the T. gondii Family A-E proteinswere identified to be largely merozoite stage-regulated(Additional file 2: Figure S1). Family A is predominantlymerozoite-specific, with 29 of 33 members expressed abovethe ≥ 8-fold abundance threshold. Only one Family A genewas expressed exclusively in tachyzoites (Additional file 2:Figure S1A and Additional file 1: Table S2). Members ofthe remaining four families (with the exception ofFamily E) were likewise strongly expressed in merozoites(>4-fold change to tachyzoites), but just below the thresh-old (Additional file 2: Figure S1B). The “miscellaneous”category (Figure 2) also contains members of additionalstructural protein families many of whose members arestrongly stage-regulated. One example is the Lysine-Arginine Rich Unidentified Function (KRUF) family of

proteins [10] (Additional file 1: Table S3). Twelve of the14 KRUF family members are strongly expressed in tachy-zoites but not expressed (≥8-fold lower) in merozoites.

Genome annotationRNA-Seq data provides valuable information about theaccuracy of existing gene models in ToxoDB [11]. In thiscase the newly generated datasets complement existingRNA-Seq data, currently from tachyzoite, bradyzoite,and oocyst stages. To estimate how many gene modelshave no associated RNA-Seq data we used ToxoDBsearch strategies. While a detailed revision of ToxoDBgene models based on RNA-Seq data will have to awaitadditional data from cat-derived gametocytes, we usedsimple ToxoDB in-silico search strategies with stringentcriteria to identify merozoite-specific gene models thatare currently without any evidence for expression. Severalexamples could be discovered by collating all predictedgene models for which no or insufficient (<40th percentile)evidence for RNA (n = 42) or protein expression (n = 38)exist at any stage of development and comparing themwith the datasets developed in this study. We eliminatedany entries with RPKM values >10 in tachyzoites and <10in merozoites resulting in 10 genes for which significantexpression is only documented in merozoites. The groupcomprises 6 Family A genes (Additional file 2: Figure S1),three hypothetical genes and SRS15B. Although suchsearch strategies are difficult to threshold and require

Figure 3 Differential expression of SRSs and genes coding for secretory organelle (microneme, rhoptry, dense granule) proteins.Bar graphs indicate –fold difference in mRNA levels (DESeq mapped reads); red (tachyzoite), green (merozoite).

Hehl et al. BMC Genomics (2015) 16:66 Page 5 of 16

manual follow-up and verification, they represent exam-ples of data mining efforts which are well within reach forexperienced ToxoDB users. In addition, fortuitous discov-ery of large numbers of merozoite-specific RNA-Seq readsmapping to regions without predicted gene models, e.g. ina 4 kb region between TGME49_297820 and TGME49_297830 on chromosome II, indicates that systematic map-ping of merozoite data will lead to discovery and annota-tion of many novel protein coding genes.

Metabolic capacity is similar in merozoites andtachyzoitesMerozoite populations expand rapidly by asexual divisionwhilst remaining strictly confined to the enterocyte mono-layer (Figure 1A, B) and presumably are confronted withless variability in terms of bioavailability of metabolitesand building blocks than tachyzoites which infect differentcell types. In addition, the parasite is ideally positioned tointercept the stream of nutrients flowing from the apicalbrush border to the basolateral face of the host enterocyte.In the absence of data on enzyme activity levels, mRNAexpression can be used effectively to model metabolic fluxdistributions and stage-specific changes [12,13]. Globalcomparative analysis of metabolic gene expression wasdone on the subset of genes used for constructing theiCS382 metabolic model for Toxoplasma [14]. Although

we identified 11 “tachyzoite-specific” and 7 “merozoite-specific” genes based on >8-fold difference in mRNA ex-pression within this subset (Additional file 2: Figure S2),there is little evidence in our data that any of these wereclustered either on chromosomes or in any particularmetabolic pathway. Nevertheless, several genes of the en-ergy metabolism are strongly regulated, with the glycolyticenzyme enolase 1 (ENO1, TGME49_268860) showing thehighest difference in expression, albeit with relatively lowRPKM values, and is effectively silenced in merozoites(RPKM 80 in tachyzoites, vs 1 in merozoites, adjusted pvalue: 1.74e-7). ENO1 is a bradyzoite marker and thus notwell expressed in tachyzoites to begin with. Indeed, ENO1appears to be dispensable in tachyzoites: cells with a tar-geted deletion of the gene grow well in vitro and in vivo,but cyst numbers in chronically infected mice are signifi-cantly decreased [15]. Conversely, the ENO2 (a canonicaltachyzoite marker) gene is strongly downregulated inbradyzoites. In merozoites ENO2 mRNA is detectedat ~2-fold higher levels (RPKM 620) than in tachyzoites,which points to an adaptive response geared towards ahigher glycolytic throughput compared to tachyzoites. Al-though a more detailed analysis of energy metabolismwould require RNA-Seq data for these enzymes in all coc-cidial stages and also bradyzoites integrated into a fluxmodel such as iCS382 [14], it is tempting to speculate that

Hehl et al. BMC Genomics (2015) 16:66 Page 6 of 16

these adaptations in merozoites help sustain rapid growthin conditions of low oxygen tension in the gut. In additionto their function in glycolysis, nuclear targeted ENO iso-enzymes were shown bind to chromosomal DNA andmodulate gene expression [15]. The significance of lowlevels of ENO1 expression in the tachyzoite dataset andthe tight silencing of the gene in merozoites in the contextof their regulatory functions remain to be investigated. Thestage-specific regulation of ENO expression is also consist-ent with 3.5 fold higher levels of phosphoglycerate kinasePGKI (TGME49_318230) RNA-Seq reads in merozoites. Inline with this, differential expression of the two glycolyticlactate dehydrogenase isoenzymes LDH1 and LDH2 [16] inmerozoites is very similar to tachyzoites but even more pro-nounced: virtual silencing of LDH2 (RPKM 2) but strongexpression of LDH1 (RPKM 962 vs 472 in tachyzoites).There are also several examples for tachyzoite-specific ex-pression of metabolic factors: both NADPH-generatingglucose-6-phosphate 1-dehydrogenase (TGME49_294200),an enzyme of the pentose phosphate pathway, and 8-amino-7-oxononanoate synthase (TGME49_290970), a keyenzyme in the biotin synthesis pathway, are not expressedin merozoites (RPKMs 5 and 1, respectively). Biotin pro-vides the prosthetic group for carboxyl transferases andhas essential functions in many anabolic and catabolic re-actions. Because biotin is contained in food but also sup-plied by gut microorganisms, a complete shutdown of itssynthesis in merozoites suggests that developmental stagesin enterocytes rely entirely on scavenging the easily access-ible protein.The most strongly regulated metabolic gene in mero-

zoites relative to tachyzoites is glycerol-3-phosphate de-hydrogenase (NAD+) (TGME49_210260), with essentiallyno expression in the latter (RPKM 12) but a 20-fold higherexpression in merozoites (adjusted p-value: 2.7e-21)(Additional file 2: Figure S2). RNA-Seq reads for thehomoserine O-acetyltransferase gene coding for an en-zyme of the methionine biosynthesis pathway are similarlyincreased in merozoites. Interestingly, expression of twogenes involved in purine metabolism, adenylosuccinatesynthetase, (a.k.a., IMP-aspartate ligase, TGME49_279450),and hypoxanthine-xanthine-guanine phosphoribosyl trans-ferase (HXGPRT, TGME49_200320) is > 8-fold higher inmerozoites. The latter, a key enzyme in the purine salvagepathway [17], is expressed at a very high level in merozoiteswith an RPKM value of 882 compared to 106 in tachy-zoites. Another purine metabolism gene, purine nucleosidephosphorylase (PNP, TGME49_307030), is expressed 6-fold higher in merozoites (RPKM 88, adjusted p-value:1.9e-2), suggesting a possible expansion of purine salvagethroughput at least during the asexual phase of parasitepopulation growth during enteroepithelial development.Altogether, we conclude that in contrast to the marked

differences in gene expression of secreted proteins in

tachyzoites and merozoites, the expression of genesinvolved in metabolic processes is largely unaffected.Hence, metabolism is predicted to be broadly similarbetween tachyzoites and merozoites, although there maybe some fine-tuning in the latter for optimal growth inenterocytes. However, no fundamental differences weredetectable that would indicate a radically different envir-onment or nutrient availability for growth in enterocytes.A more detailed examination and testing of specific hy-potheses using for example the recently developed fluxbalance analysis models to identify stage-specific meta-bolic bottlenecks [14] will be required to confirm orrefute this current interpretation as well as identify po-tential drug targets.

SRS proteins are widely expressed in merozoitesThe T. gondii genome encodes several distinct, coccidian-specific surface gene families, including the SRS and SAG-unrelated surface antigens (SUSA) [18]. Arguably the moststrikingly regulated set of genes in the merozoite datasetwere those coding for the 111 members of the SRS super-family of proteins annotated in ToxoDB (Version 8.1);expression of more than 52 members of this family waspresent in merozoites, whereas a separate set of 14 SRSgenes were expressed exclusively in tachyzoites (Figure 3).Nearly half of the merozoite expressed genes coding forSRS proteins were present in 5 clusters, SRS12, 15, 22, 26and 55. SRS proteins are involved in attachment to hostcells, but also provoke immune reactions and regulateparasite virulence, which is thought to promote the forma-tion of tissue cysts in intermediate hosts in order to estab-lish persistent, latent infections that facilitate transmissionof infection to the definitive host. Previous work hasshowed that tachyzoites differentially express a number ofSRS genes [18], and these expression differences have beenpostulated to account for the ability of this stage to invadea broader range of host cells than other coccidians [10].However, merozoites, which only infect a single cell type(the feline enterocyte) co-dominantly expressed a largerepertoire of 52 SRS proteins in a developmental life-cyclestage-dependent manner (Additional file 2: Figure S3). Al-ternatively, recent work has established that the SRS foldis present in the 10 member Pfs-230-related 6-Cys familyof Plasmodium adhesins [19] that facilitate gamete-gamete recognition and promote gamete recognitionand fertilization. This may suggest that the merozoite-restricted SRS genes are less relevant for attachment andinvasion within enterocytes, but rather promote gametedevelopment and fertilization. Given the demonstratedability of SRS proteins to promote immune responses, it isalso conceivable that they play a role in stimulating intes-tinal inflammation and diarrhea, to facilitate the produc-tion and dispersal of oocysts.

Hehl et al. BMC Genomics (2015) 16:66 Page 7 of 16

The striking stage-specificity of expression of SRSgenes in merozoites versus tachyzoites (and as previouslyobserved for merozoites versus bradyzoites) raises ques-tions about how their expression is regulated. One clueto this can be obtained from analysis of chromosomaldistribution of the various SRS genes [10,18]. However, wefind little evidence of concerted clustering of merozoite-or tachyzoite-specific SRS genes on distinct chromosomes(Figure 4), which is not a complete surprise since T. gondiiSRS genes are dispersed across all chromosomes and notexclusively in a subtelomeric regional distribution patternfor surface antigen genes that is characteristic for otherprotozoa [20-23]. Thus, rather than relying on sequentialexpression of a single gene or promoting rapid ectopic

Figure 4 Chromosomal distribution and relative expression levels ofposition and distribution of SRS (◊), GRA (⎔) and T. gondii protein A and Dtachyzoites is displayed. Tandemly repeated genes are shown as clusters. Unco(RPKM <10). Black coloured genes were not differentially expressed between tarelative to merozoites) or green (induced in merozoites relative to tachyzoitesstage. The chromosomal position of SRS pseudogenes is not displayed. The mexpression was specifically induced in merozoites. The majority of SRSs were uwhereas only 14 were upregulated in tachyzoites. Genes in each cluster tendeonly 4 exceptions: One gene in each of the SRS16, SRS36, T. gondii Family A (Cupregulated in tachyzoites, relative to merozoites.

recombination rates to generate surface antigenic varia-tion, T. gondii expresses relatively large sets of non-overlapping, co-dominant SRS transcripts [18,24].

Merozoites express a distinct set of microneme proteinsMicronemes are small organelles clustered at the anter-ior end of apicomplexan parasites and generally secreteadhesive proteins that are important for motility andinvasion. Microneme (MIC) proteins contain one or se-veral adhesive domains such as thrombospondin, vonWillebrand factor A, epidermal growth factor (EGF) orPAN-domains (reviewed in [25]). Comparative analysis ofRNA-Seq data shows that the majority of the 26 Toxo-plasma MICs and MIC2 associated protein (MIC2AP)

SRS, GRA, and T. gondii protein A and D families. The chromosomalfamilies (Δ) that are differentially expressed between merozoites andloured genes were not expressed in either tachyzoite or merozoite >stagechyzoites and merozoites. The shade of red (induced in tachyzoites) indicated the fold increase in expression relative to the other life cycleajority of GRA genes were upregulated in tachyzoites. Only GRA11 genepregulated in merozoites. 52 SRS genes were upregulated in merozoitesd to be coordinately regulated according to life cycle stage, withhromosome XII) and T. gondii Family D (Chromosome XI) clusters was

Hehl et al. BMC Genomics (2015) 16:66 Page 8 of 16

accumulate a high number of mapped tachyzoite reads(Additional file 2: Figure S4). Conversely, the range ofMIC genes that are specifically expressed in merozoites isconsiderably reduced and restricted to the highly ex-pressed genes coding for PAN-domain containing proteinsMIC17A-C (Additional file 2: Figures S4 and S5), MIC12,a gene annotated as a putative MIC, and a microneme-like protein. Interestingly, both MIC2 and MIC2AP [26],as well as MIC6-1-4 [27], which are strongly expressed intachyzoites and secreted as functional complexes, arevirtually silenced in merozoites (Figure 5). Similarly, theRNA-Seq reads for the MIC3 escorter MIC8 [28] are>100-fold lower in merozoites, whilst MIC3 mRNAs areonly 6-fold less abundant. Since trafficking to the organ-elles, as well as deployment on the surface, seems to occurin complexes as a rule [29], we analyzed the predictedtopologies of the six merozoite-specific MIC proteins.None had the predicted transmembrane domain and cyto-plasmic tail typical for canonical escorters such as MIC6[30]. In tachyzoites these adhesin complexes are essentialcomponents of the molecular motor and play a key role inproviding the machinery for gliding motility [31].The number of microneme organelles of different

Apicomplexa has been correlated with the requirementfor gliding motility [25], ranging from many in Eimeriasporozoites and merozoites [32] to none in Theileriazoites [33]. Toxoplasma tachyzoites, which display vigor-ous gliding motility and are able to cross cell barriers, have

Figure 5 Differential expression of three MIC escorter/adhesin complfor MIC escorter proteins.

many microneme organelles and express a wide range ofMIC proteins. Although T. gondii merozoites have a highnumber of microneme organelles [5], the complement ofexpressed MIC and other genes (e.g. AMA1 [34]) codingfor proteins secreted from micronemes appears to besignificantly restricted, suggesting that the size of theexpressed protein complement is not necessarily corre-lated to the number of organelles present in the cell. Onthe other hand, the finding that prominent tachyzoiteMIC complexes are not expressed in merozoites(Figures 3 and 5), could be explained by fundamentaldifferences in their mode of gliding motility and cellinvasion in contrast to the extraintestinal developmen-tal stages. Because merozoite proliferation is restrictedto enterocytes, the most obvious difference is that ma-ture merozoites egress into the gut lumen and invadenew host enterocytes from the exposed apical side. Thisrequires some migration through intestinal contentsand within the mucus layer rather than through inter-cellular spaces and on cell surfaces. Because freshly iso-lated extracellular merozoites display vigorous glidingmotility (data not shown) it is unlikely that these stagesare without surface adhesins linked to the glideosomecomplex [31,35]. Rather, as yet unknown MIC adhesinsspecific for enterocyte surface receptors and containingnon-canonical adhesive domains will likely be discov-ered among hypothetical proteins that are expressed ina merozoite stage-specific manner.

exes. Expression values are indicated as [RPKM], asterisks: genes coding

Hehl et al. BMC Genomics (2015) 16:66 Page 9 of 16

Merozoites only weakly express genes of the tachyzoitemoving junction complexThere is a striking difference in expression of genescoding for rhoptry neck proteins (RONs) and AMA1 intachyzoites versus merozoites (Figure 6). In particular,mRNA levels of genes coding for proteins that combineto form the moving junction as part of the tachyzoite in-vasion machinery (i.e., AMA1, RON2, RON4, RON 5and RON8; Figure 6) [36,37] are dramatically lower inmerozoites. This supports the contention, raised above,that the mechanism of gliding motility and by extensioncell invasion is distinct in merozoites. The dramaticallylower expression of the canonical tachyzoite AMA1(TGME49_255260) in merozoites is not compensated bya significant increase in expression of the sporozoite-specific AMA1 paralog and the associated sporoRON2[38] is not expressed. A third Toxoplasma AMA1 para-log (TGME49_300130) is weakly expressed in bothstages, and its functional role in invasion, if any, remainsto be demonstrated. Interestingly, previous claims thatAMA1 is essential for tachyzoite invasion [39] has beenrevised in light of recent data showing that AMA1knockout tachyzoites invade host cells normally in vitro[40]. On the one hand, gene expression data alone pre-dicts that invasion via the massively expanded surface ofthe enterocyte brush border by merozoites is distinctfrom the well described process observed in extrain-testinal stages. On the other, the exact requirements andconcepts for formation of a tight junction [36,41-44]which will be moved towards the posterior end of the in-vading parasite by motor proteins are currently beingrevisited [40,45]. While gene expression data indicatesfundamental differences in cell invasion by merozoites,the question whether merozoites assemble an alternative

Figure 6 Genes coding for components the tachyzoite moving junctioin merozoites. Cartoon depicting the RON and AMA protein assembly at(RPKM values) of components and paralogs in tachyzoites and merozoites.

moving junction machinery for invasion or invade en-terocytes by an entirely different mechanism, e.g. basedon membrane-zippering such as in Theileria [46] orNeospora [47], requires further investigation.

Merozoites express a distinct set of rhoptry proteinsWe also investigated RNA expression for a group of 57genes coding for rhoptry proteins (ROPs) compiled fromannotated gene models on ToxoDB (Version 8.1) [48]and expanded by manual annotation (Additional file 1:Table S4). In total 24 ROP genes were significantlydifferentially expressed (≥8-fold abundance threshold) bet-ween tachyzoites and merozoites (Figure 3 and Additionalfile 2: Figure S6); 19 genes were considered not expressedin merozoites (RPKM values were <10 in all cases) whereastwo were expressed at very low levels (RPKM< 40) in mer-ozoites. Four ROP genes were, however, expressed speci-fically in merozoites. Interestingly, in a survey of knownfunctions of ROP proteins, we find that the prominentmurine virulence factors ROP16 and ROP18/5 are notexpressed in merozoites.ROP16 and 18, which both possess a conserved cata-

lytic triad and phosphoryltransfer activity, have beenidentified as secreted, strain-specific virulence factors inmice. ROP16 phosphorylates the signal transducers andactivators of transcription (STATs) 3 and 6, and re-gulates interleukin 12 levels [49,50]. The polymorphicROP18 kinase phosphorylates and thus inhibits loadingof immunity-related GTPase (IRG) 6 onto the PV mem-brane, thereby preventing parasite clearance in infectedmice by activated macrophages [51,52]. ROP5, a pseudo-kinase with an altered catalytic triad, regulates the kinaseactivity of ROP18 [53] possibly by an allosteric mecha-nism. Therefore, the absence of IRGs in cats is the

n invasion machinery are expressed at significantly lower levelsthe moving junction complex (left); bar graph showing expressionMt, microtubules; PM, host cell plasma membrane.

Hehl et al. BMC Genomics (2015) 16:66 Page 10 of 16

simplest explanation for the silencing of the ROP18 (andpossibly the ROP5) gene in merozoites rendering expres-sion and secretion of these ROPs as part of a defensestrategy redundant during enteroepithelial development.

Merozoites express a distinct set of dense granuleproteinsWe detected significant differences in expression withinthe current set of 25 annotated genes encoding densegranule (GRA) proteins (Figure 3, Additional file 2:Figure S7). There was a significant overall bias towardshigh expression in tachyzoites. Twelve GRA genes werenot expressed in merozoites (RPKM < 10), whilst only 2GRA genes (GRA11 isoforms TGME49_212410 andTGME49_237800, which have, as yet, undefined func-tions) are >8-fold higher expressed in cat stages (18- and13-fold, respectively; Figure 3). The high expression ofGRA22 in tachyzoites is intriguing, since this proteinregulates parasite egress from host cells [54] and the ab-sence of GRA22 expression in merozoites is thereforepuzzling. This might imply that egress from cat entero-cytes is less coordinated by the parasite and this maypartially explain why parasite development is asynchron-ous in the cat intestine. GRA genes expressed in tachy-zoites fall into two main groups: moderately highlyexpressed GRA genes (RPKM values in the range of 100)code for proteins that are exported across the PV mem-brane that function to alter host gene expression (i.e.,GRA 15, 16, and 24); versus highly expressed GRA genes(RPKMs > 103) some of whose products participate inthe formation of the tubulovesicular network (TVN)within the PV space. Our RNASeq data shows thatGRA15 is not expressed in merozoites. This tachyzoitevirulence factor has a well-defined role in the activationof several host cell pathways that trigger potent pro-inflammatory responses causally associated with tachy-zoite to bradyzoite conversion [49,55-59]. In mice, parasitesurvival is dependent on combined expression of GRA15with the STAT-activating form of ROP16 [49]. BothGRA15 and ROP16 genes are not expressed in merozoites,strongly suggesting that JAK-STAT3/6 and NF-κB sig-naling pathways in enterocytes of the definitive host arenot targeted by these secreted factors. GRA24 and 16 arealso not expressed in merozoites. GRA24 was recentlyidentified as a parasite soluble effector that traffics to thehost cell nucleus, interacts with p38, and is thought tosynergize with GRA15 to promote production of the pro-inflammatory cytokine, IL12 [60]. Similarly, GRA16 isexported to the host cell nucleus and affects transcriptionof host cell genes involved in metabolism and cell cycleprocesses [61]. Whether these host cellular processes arenot targeted by merozoite gene products during theirasexual expansion in cat enterocytes, or if as yet

unannotated merozoite-specific paralogs exist to performthis function, remain to be established.The RPKM values for GRA2, 4, and 6 genes, coding

for three proteins which are directly associated with theTVN in the PV [62,63], are >8-fold lower in merozoitescompared to tachyzoites (Additional file 2: Figure S7).The silencing of these GRAs (RPKMs <10 and log [2]-fold change ≥ 8) in merozoites is consistent with electronmicroscopy data [5] showing that type C-E schizonts incat enterocytes do not elaborate the extensive TVNwhich is typical of vacuolar spaces of tachyzoites, brady-zoites and type A and B schizonts. This suggests thatlater stage schizonts dispense with the TVN because oftheir different mode of cell division [5]; parasites in typeC-E schizonts multiply by endopolygeny rather than byendodyogeny as early stage merozoites, tachyzoites andbradyzoites. Magno et al. [64] suggested that an im-portant structural function of the TVN might beorganization of daughter cells within the PV duringmultiplication. Involution of the TVN and repression ofgenes coding for factors associated with these mem-branes in merozoites is consistent with endopolygenyand attachment of parasites to residual bodies [5]. Con-versely, the GRA7 gene, whose 36 kDa product is involvedin nutrient acquisition by trapping host endosomal-lysosomal vesicles in the PV [65], is expressed at ~4-foldlower levels in merozoites (Additional file 2: Figure S7)but is still relatively abundant with an RPKM value of 351(1837 in tachyzoites). Thus, although not always asso-ciated with the PV membrane, GRA7 appears to be uni-versally expressed throughout the life cycle [66].Additional factors secreted by dense granules include

Kazal-type protease inhibitors PI1, PI2 (TgME49_217430,TgME49_208450) [67-69] (Table S5), which are secretedinto the PV and appear to be differentially regulated, des-pite strong conservation on the structural and sequencelevel. Whilst read numbers for the PI1 gene, the mostabundantly expressed protease inhibitor in tachyzoites[69], are >2-fold higher in merozoites, PI2 is not expressedin merozoites (RPKMs 50-fold lower). Three of five ad-ditional genes annotated as Kazal-type family protease in-hibitors (Table S5) show significantly higher expression inmerozoites. However, whether the four predicted secretedmembers of this group also traffic via dense granules isunknown.

Trans and cis regulators of merozoite and tachyzoitebiologyBoth trans and cis regulators are known to play crucialroles in the transcription of stage specific genes in T. gondii.The trans-acting ApiAP2 transcription factors were dis-covered in the Apicomplexa almost 10 years ago [70] andregulate various developmental processes in T. gondii, in-cluding bradyzoite conversion, with TgAP2XI-4 [71], and

Hehl et al. BMC Genomics (2015) 16:66 Page 11 of 16

TgAP2IX-9 [72] promoting and restricting tissue cystformation, respectively, as well as the tachyzoite cell cycle,with TgAP2XI-5 potentially regulating hundreds of cellcycle-dependent genes [73]. Using stringent criteria weidentified three significantly merozoite- or tachyzoite-specifically expressed ApiAP2-coding genes in our study(Additional file 2: Figure S8). One merozoite-specific geneTgAP2VIIa-I, and two tachyzoite-specific genes, TgAP2-XII-I and TgAP2VIIa-II showed >8-fold difference in theirexpression levels, suggesting that these transcription fac-tors may play a role in regulating a suite of stage-specificgenes. Relaxing the criterion to include genes differing ≥4-fold in mRNA abundance identified 8 additionaltachyzoite-, and 2 additional merozoite-specific genes (atotal of 13 regulated stage-specific TgAP2s). Interestingly,several other TgAP2 genes were expressed at very lowlevels, and in the case of TgAP2XII-3, can be considered si-lenced in both stages. Overall, differential expression ofTgAP2 is not extensive between tachyzoite and merozoites,and this may reflect their similar life cycle fate as rapidlyamplifying asexual stages. However, the evenly graded dis-tribution of RPKM values (Additional file 2: Figure S8B)suggests that some differences may become more pro-nounced when merozoites differentiate to gametocytes.Moreover, the two datasets compared here do not repre-sent synchronized populations, which would average outany dynamics occurring during the merogony cell cyclespecifically during development from stage A throughstage E schizonts. By contrast, 24 of the 67 predictedTgAP2-coding genes were found to be differentiallyexpressed at different check-points of the tachyzoite cellcycle in a microarray analysis of synchronized parasites[74]. In addition, ApiAP2s have been shown to bind othertranscriptional regulators such as histone lysine acetyl-transferases [75], histone deacetlyase [76] and even otherApiAP2s [77], producing interactions that may confoundcorrelations between transcription factor activities andtheir cognate gene transcript levels.Next, we used the Regulatory Sequence Analysis Tools

on-line computational resource (http://www.rsat.eu/) toperform a global analysis of upstream flanking sequencesof stage-specifically expressed genes. The analysis returnedseveral enriched six-base DNA motifs within the predictedpromoters, i.e. 500 bp genomic regions upstream ofthe transcriptional start site of either merozoite- ortachyzoite-specific genes. Listed in Figure 7A are the fivemost significantly enriched 6 bp motifs found. Patternmatching between these enriched motifs produced anamalgamated seven-base DNA motif in both merozoite-and tachyzoite-specific gene promoters, representingputative cis-regulatory elements (Figure 7B and C, re-spectively). The GAAGAAA motif, present in 21.5% ofmerozoite promoters, was also present in 19.2% of tachy-zoite promoters, indicating that this motif is unlikely to

represent a merozoite-specific cis-regulatory element(Figure 7D). However, the GAGACGC motif is clearlyenriched in the promoters of tachyzoite-specific genes(present in 20.4% of promoters) over merozoite-specificgenes (6.5%). The distribution of this putative cis-regula-tory element in the predicted promoters of the 15 mosttachyzoite-specific genes is illustrated in Figure 7E. Thismotif is nearly identical to a motif (A/TGAGACG) pre-viously identified as a functional cis-element in the pro-moters of dense granule genes, such as TgGRA1, TgGRA2,TgGRA5 and TgGRA6, as well as the tachyzoite marker,TgSRS29B (aka TgSAG1) [78]. A sequence-specific tran-scription factor capable of activating gene transcriptionvia this cis element has yet to be identified; however, itmay play a significant role in promoting and/or maintain-ing a tachyzoite developmental state.

ConclusionsTaken together, our data support a correlation betweenpromiscuity of invasion and dissemination of the T. gon-dii parasite in a variety of host cells and tissues. Inaddition, the data presented here underscore the phe-nomenon of non-overlapping stage-specific expressionof gene sets throughout development (e.g. SRS, ROP,GRA, MIC) as previously proposed [79] within the Toxo-plasma life cycle [18]. Hence, the differentiation of bra-dyzoites into merozoites, which occurs only in the felineintestine, entails the shut down or downregulation ofsuites of tachyzoite-specific genes that would normallybe activated to support Toxoplama’s promiscuous abilityto invade and reproduce asexually in virtually any nucle-ated cell from essentially any warm-blooded animal. Thisstudy provides compelling datasets to begin to explorethe suites of genes activated to support merozoite inva-sion, differentiation, and expansion in cat enterocytesand establishes T. gondii as an excellent genetically-amenable model system to study schizont asexual ampli-fication prior to gamete development, the hallmark ofcoccidian development in the definitive hosts thesecoccidian parasites infect. Our data suggest that glidingmotility and the formation of a moving junction (or analternative invasion apparatus) requires machinery thatis specifically adapted to breaching the enterocyte sur-face barrier for every cycle of replication, egress and inva-sion. Studies of the SRS proteins exclusively expressedduring coccidian development will determine if theseproteins promote attachment and invasion of feline enter-ocytes, or whether they participate in gamete-gameterecognition to promote fertilization, analogous to theorthologous role performed by related 6-Cys family pro-teins during Plasmodium development [80]. Understan-ding the governance of gene expression and regulationwithin these families, e.g. via interactions with the TgAP2family of transcriptional regulators, may prove key to

Figure 7 Identification of cis regulatory elements in the predicted promoters of merozoite- and tachyzoite-specific genes. (A) Enrichedsix-base motifs were identified within the predicted promoters of merozoite- and tachyzoite-specific genes. Listed for both stages are the fivemost significantly enriched motifs, their occurrence, expected occurrence, and occurrence significance. Asterisks indicate motifs overlapping withthe putative cis regulatory elements for promoters of merozoite- and tachyzoite-specific genes, shown in (B) and (C), respectively. (D) Percentageof promoters from merozoite- or tachyzoite-specific genes that contain either the GAAGAAA or GAGACGC putative cis elements. (E) Distributionof the GAGACGC cis element (blue) within the promoters of the fifteen most tachyzoite-specific genes. The predicted promoters correspond tothe 500 bp genomic region directly upstream of the transcriptional start site, indicated by an arrow.

Hehl et al. BMC Genomics (2015) 16:66 Page 12 of 16

dissecting the full panorama of pathways that regulate thedefinitive coccidian lifecycle and aid in the complete andaccurate annotation of the T. gondii genome database.

MethodsIsolation of enteroepithelial parasite stages from a catinfected with bradyzoites from sheepOocysts of the CZ T. gondii isolate (type II) were origi-nally isolated from the feces of a captive Siberian tiger(Panthera tigris altaica) at the Dvůr Králové Zoo (CzechRepublic). The strain isolation was performed in 2005 byDr. B. Koudela, and the strain was maintained by passagesbetween mice and cats [81,82]. After obtaining the strainfrom Dr. Koudela the parasites were passed once throughmice and cats to generate ~50 million CZ oocysts, whichwere used to infect four sheep (104 oocysts per sheep).Equal amounts of brain tissue containing bradyzoites intissue cysts from three chronically infected sheep washomogenized and fed to cats (55 g per cat). A semi-

quantitative analysis of parasite burden in sheep braintissue was done by magnetic capture real time PCR, toconfirm that equal quantities of parasites were inoculatedinto each cat but no definitive count of cysts was per-formed. Pilot experiments had established that this proto-col results in strong and reproducible infections of cats.All feces from experimentally infected cats were collectedand oocysts were enumerated by microscopy every day.For isolation of enteroepithelial stages of T. gondii a catthat was experimentally infected with brain cysts from asheep was euthanized after onset of patency (day 5 post in-fection). The small intestine was clamped, removed, andimmediately cooled to 0°C. All subsequent procedures forparasite isolation and purification were performed in acold room (4°C) and on ice. The small intestine (~70 cm)was carefully rinsed with PBS, divided into four sections ofequal length, and opened. Enterocytes were selectively re-moved from villi by gentle scraping with a rubber police-man (Figure 1B). Tissue samples for histology were taken

Hehl et al. BMC Genomics (2015) 16:66 Page 13 of 16

before and after removal of enterocytes. The cells contain-ing numerous parasites were washed off with ice cold PBSand collected by centrifugation (10 min, 900 × g, 2°C). En-terocyte preparations were evaluated microscopically afterfixation and staining to identify parasite stages. The twoenterocyte samples collected from two different sectionsof the small intestine were washed with buffer containingTween80 and syringe-passaged The samples appeared toconsist exclusively of free merozoites as no other stages(i.e. gamonts) could be identified by microscopy (Figure 1C).Histological and fluorescence microscopy examination

of small intestinal tissue sections revealed numerousparasite stages that were readily identified as merozoites,schizonts (Figure 1A) and also occasional gamonts dueto their very distinctive morphologies. However, the finalpreparation was clearly highly enriched for merozoitesallowing us to focus on interrogating RNA expressiondatasets specifically for differences between asexually de-veloping merozoites and tachyzoites. The enrichment formerozoites or, more precisely, the paucity of gameto-cytes in the two biological replicates used for RNA-Seq,was indeed borne out by a lack of known gametocytemarkers (e.g. genes coding for flagellar or oocyst wallproteins) in the RNA-Seq datasets. This is in contrastwith a preliminary gene expression analysis of stages iso-lated at or past peak patency (Ramakrishnan et al. un-published). CZ strain tachyzoites were grown in vitro;confluent human foreskin fibroblast (HFF) cells were in-fected at an MOI (>5) and harvested from synchronizedcultures when 1/3 of the vacuoles had spontaneouslylysed.

Ethics statementExperiments involving animals were performed underthe direct supervision of a veterinary specialist, and ac-cording to Swiss law and guidelines on Animal Welfareand the specific regulations of the Canton of Zurich.Permit number 108/2010 covers all animal experimentspresented in this paper and was approved by the Veter-inary Office and the Ethics Committee of the Canton ofZurich (Kantonales Veterinäramt Zürich, Zollstrasse 20,8090 Zürich, Switzerland).

RNA isolation and quality controlInfected tissue and parasite samples were resuspendedin 600 μl buffer RLT (QIAGEN) containing 10 μl/mlβ-mercaptoethanol. Cell suspensions were then passedthrough a QIAshredder (QIAGEN) column by centrifu-gation at ≥ 8000 g for 1 min. The RNA was then purifiedusing the RNeasy Mini Kit (QIAGEN) according to themanufacturer’s protocol (including an on-column DNAdigest), and eluted in RNase-free water. The quality of theRNA was analysed using the Agilent RNA 6000 Pico Kit(Agilent) and a Bioanalyzer 2100 (Agilent) (Figure 1D).

RNA concentration was determined using a Qubitfluorometer (Invitrogen) together with the RNA assay(Invitrogen).The genome-wide transcriptome library, was produced

using the MicroPolyA Purist Kit (Ambion) and theSOLiD Total RNA-Seq kit (Applied Biosystems). Briefly,approximately 300-500 ng of mRNA were enrichedstarting from 15-20 μg of total RNA, using MicroPolyAPurist Kit (two rounds of purification to receive onlyminor ribosomal RNA contamination). The quality andthe quantity of the extracted polyA RNA was assessedusing a Bioanalyzer (Agilent) Picochip and a Qubitfluorometer (Invitrogen), respectively. The mRNA wasthen fragmented using RNase III. Ligation of the adaptormix and reverse transcription were performed followingthe manufacturer’s protocol. cDNA libraries were se-lected for fragment sizes between 150 and 250 bp, amp-lified for 15-18 cycles of PCR using barcoded adaptorprimers and purified using the PureLink PCR micro kit(Invitrogen). Library size and concentration were assessedusing a Bioanalyzer (Agilent) and a Qubit fluorometer(Invitrogen), respectively.The poly-A transcriptome libraries were used for

emulsion PCR (e-PCR) using a concentration of 0.5 pM.The barcoded libraries were pooled before e-PCR andthe resultant beads were loaded on a full SOLiD 5 slide(Applied Biosystems), according to manufacturer’s in-structions. SOLiD ToP Sequencing chemistry was usedto produce paired-end (50 bp + 35 bp) sequencing reads.

Comparative transcriptomics of merozoites andtachyzoitesWe generated RNA-Seq datasets from CZ strain enter-oepithelial parasites isolated from mechanically strippedenterocytes and from tachyzoites cultured in humanforeskin fibroblasts on a SOLiD5 platform (Applied Bio-systems). Two biological samples were sequenced for eachstage (two separate parasite and RNA preparations fromdifferent sections of the small intestine as well as two dif-ferent tachyzoite cultures). The samples contained diffe-rent proportions of parasite and host RNA (Figure 1D).Adjustments to account for these differences were: i) re-moval of all reads mapping to tRNAs, and ribosomalRNAs prior to DESeq analysis, and ii) normalizationand weighting by using transcripts lengths to calculateRPKMs. The reads in colorspace were aligned to the T.gondii ME49 genome from ToxoDB v8.1 using SHRiMP2mapper [83,84].The count table of reads fitting the genecoordinates have been processed by Bioconductor libraryDESeq [8,85] to produce the table that includes the meanread number for each gene together with a p-value andfold change. We have applied an 8-fold (log2 ≥ [3]) changeas a cut-off for differential expression.

Hehl et al. BMC Genomics (2015) 16:66 Page 14 of 16

Immunofluorescence assay and microscopySamples from the small intestine of infected cats were em-bedded in paraffin. Sections of 3 μm were placed on slideswith a positively charged coat. After drying over night at37°C, slides were deparaffinized 3 times 2 min in xyleneand then washed twice 1 min in 100% ethanol. Slides werethen washed 1 min each in 96%, 70% ethanol, and water.Slides were then submerged in alkaline Target RetrievalSolution pH 9 (Dako), boiled for 20 min in a pressurecooker at 96°C and transferred to water. Permeabilizationwas done in PBS/0.3% Triton X-100 (permeabilizationbuffer, PB) for 5 min. Blocking was performed in rabbitserum for at least 1 h at RT. After two washes in PB for5 min, sheep immune serum against T. gondii diluted in20% rabbit serum/PB was used to label the parasites byovernight incubation at 4°C. Slides were washed threetimes with PB for 5 min and samples were incubated withrabbit anti-sheep FITC in 20% rabbit serum/PB for 1 h atRT. Slides were then washed once in PB and the nucleiwere counter stained with 1 μg/ml DAPI in PB for15 min. Slides were washed again twice with PB for 5 minand the samples were mounted using Vectashield (VectorLaboratories). Imaging was performed using a Leica DMI6000 B microscope and the Leica LAS AF software. Im-ages were processed using ImageJ version 1.47.

Availability of supporting dataThe raw sequence data and a complete description of thestudy are available at the European Nucleotide Archive(Study accession no. PRJEB7935). The data are also avail-able via Toxoplasma Genomics Resource, ToxoDB, [http://www.toxodb.org/toxo/]. All other data sets supporting theresults of this article are included within the article and itsadditional files.

Additional files

Additional file 1: Table S1A. Significantly higher expressed genes inmerozoites (see also Figure 2). Table S1B. Significantly higher expressedgenes in tachyzoites (see also Figure 2). Table S2. List of annotatedT. gondii Family A genes. The color key for the expression heat map (log2fold change) is included. Table S3. Differential expression of KRUF familygenes. Table S4. List of definitively and provisionally annotated rhoptry(ROP) genes. The color key for the expression heat map (log2 foldchange) is included. Table S5. Protease inhibitors including differentiallyexpressed secreted Kazal-type proteins. The color key for the expressionheat map (log2 fold change) is included.

Additional file 2: Figure S1- S8. Differential expression of – T. gondiigenes and gene families: S1. Differential expression of – T. gondii FamilyA-E genes. S2. Differential expression of metabolic genes. S3. Differentialexpression of SRS genes. S4. Differential expression of microneme (MIC)genes annotated in ToxoDB. S5. Differential expression of genes codingfor PAN-domain containing proteins. S6. Differential expression ofannotated rhoptry (ROP) genes. S7. Differential expression of annotateddense granule (GRA) genes. S8. Differential regulation of TgAP2 genes.

Competing interestsThe authors declare that they have no competing interests.

Authors’ contributionsABH designed experiments, supervised experimental infections in vivo andin vitro, participated in generation of the biological material for analysis, andin quality controls, performed RNA sequence data analysis, directed andsupervised analysis of data, wrote the final draft of the manuscript, andedited the final figures. WUB carried out experimental infections in vivo andin vitro, carried out dissections of host animals, participated in generation ofthe biological material. CL designed experiments, carried out in vitroinfections and dissection of host animals, generated biological material foranalysis, performed quality controls, directed and supervised RNAsequencing, participated in analysis of data. CR performed RNA sequencedata analysis, directed and supervised analysis of data, designed figures andgraphs for the manuscript. MO performed RNA sequence data analysis andquality control. RAW performed RNA sequence data analysis, participated inthe preparation of manuscript drafts, designed figures and graphs for themanuscript. MEG performed RNA sequence data analysis participated in thepreparation of the final draft of the manuscript, designed figures and graphsfor the manuscript. NCS designed experiments, performed RNA sequencedata analysis, and participated in the preparation of the final draft of themanuscript. PD designed experiments, supervised experimental infectionsin vivo and dissections of host animals. All authors read and approved thefinal manuscript.

AcknowledgementsWe are grateful for the excellent technical assistance by Dr. ManuelaSchnyder, Therese Michel and Armin Rüdemann, and the support by theFunctional Genomics Center Zurich. This work was supported by fellowshipsfrom the University of Zurich to WUB and CL, and in part by the IntramuralResearch Program of the NIH and NIAID (MEG). NCS is supported by a grantfrom the Bellberry Foundation. MEG is a scholar of the Canadian Institute forAdvanced Research (CIFAR) Program for Integrated Microbial Biodiversity.RAW was supported by a Swiss Government Excellence Scholarship:Postdoctoral Scholarship from the Swiss Confederation.

Author details1Institute of Parasitology-University of Zurich, Winterthurerstrasse 266a, Zürich8057, Switzerland. 2Current address: Department of Anaesthesiology and PainMedicine, Inselspital, University of Bern, Freiburgstrasse, Bern 3010,Switzerland. 3Functional Genomics Center Zurich, Winterthurerstrasse 190,Zürich 8057, Switzerland. 4Queensland Tropical Health Alliance ResearchLaboratory, Faculty of Medicine, Health and Molecular Sciences, James CookUniversity, Cairns Campus, McGregor Road, Smithfield, QLD 4878, Australia.5Molecular Parasitology Section, Laboratory of Parasitic Diseases, NIAID, NIH,Bethesda, Maryland, USA.

Received: 5 August 2014 Accepted: 7 January 2015

References1. Halonen SK, Weiss LM. Toxoplasmosis. Handb Clin Neurol. 2013;114:125–45.2. Jones JL, Dubey JP. Foodborne toxoplasmosis. Clin Infect Dis.

2012;55(6):845–51.3. Swierzy IJ, Muhammad M, Kroll J, Abelmann A, Tenter AM, Luder CG.

Toxoplasma gondii within skeletal muscle cells: a critical interplay forfood-borne parasite transmission. Int J Parasitol. 2014;44(2):91–8.

4. Ferguson DJ. Use of molecular and ultrastructural markers to evaluate stageconversion of Toxoplasma gondii in both the intermediate and definitivehost. Int J Parasitol. 2004;34(3):347–60.

5. Speer CA, Dubey JP. Ultrastructure of early stages of infections in mice fedToxoplasma gondii oocysts. Parasitology. 1998;116(Pt 1):35–42.

6. Kim K, Weiss LM. Toxoplasma: the next 100 years. Microbes Infect.2008;10(9):978–84.

7. Torrey EF, Yolken RH. Toxoplasma oocysts as a public health problem.Trends Parasitol. 2013;29(8):380–4.

8. Anders S, Huber W. Differential expression analysis for sequence count data.Genome Biol. 2010;11(10):R106.

9. Fritz HM, Buchholz KR, Chen X, Durbin-Johnson B, Rocke DM, Conrad PA,et al. Transcriptomic analysis of toxoplasma development reveals manynovel functions and structures specific to sporozoites and oocysts. PLoSOne. 2012;7(2):e29998.

Hehl et al. BMC Genomics (2015) 16:66 Page 15 of 16

10. Reid AJ, Vermont SJ, Cotton JA, Harris D, Hill-Cawthorne GA, Konen-Waisman S, et al. Comparative genomics of the apicomplexan parasitesToxoplasma gondii and Neospora caninum: Coccidia differing in host rangeand transmission strategy. PLoS Pathog. 2012;8(3):e1002567.

11. Gajria B, Bahl A, Brestelli J, Dommer J, Fischer S, Gao X, et al. ToxoDB: anintegrated Toxoplasma gondii database resource. Nucleic Acids Res.2008;36(Database issue):D553–6.

12. Colijn C, Brandes A, Zucker J, Lun DS, Weiner B, Farhat MR, et al.Interpreting expression data with metabolic flux models: predictingMycobacterium tuberculosis mycolic acid production. PLoS Comput Biol.2009;5(8):e1000489.

13. Huthmacher C, Hoppe A, Bulik S, Holzhutter HG. Antimalarial drug targets inPlasmodium falciparum predicted by stage-specific metabolic networkanalysis. BMC Syst Biol. 2010;4:120.

14. Song C, Chiasson MA, Nursimulu N, Hung SS, Wasmuth J, Grigg ME, et al.Metabolic reconstruction identifies strain-specific regulation of virulence inToxoplasma gondii. Mol Syst Biol. 2013;9:708.

15. Mouveaux T, Oria G, Werkmeister E, Slomianny C, Fox BA, Bzik DJ, et al.Nuclear glycolytic enzyme enolase of Toxoplasma gondii functions as atranscriptional regulator. PLoS One. 2014;9(8):e105820.

16. Yang S, Parmley SF. Toxoplasma gondii expresses two distinct lactatedehydrogenase homologous genes during its life cycle in intermediatehosts. Gene. 1997;184(1):1–12.

17. Chaudhary K, Darling JA, Fohl LM, Sullivan Jr WJ, Donald RG, Pfefferkorn ER,et al. Purine salvage pathways in the apicomplexan parasite Toxoplasmagondii. J Biol Chem. 2004;279(30):31221–7.

18. Wasmuth JD, Pszenny V, Haile S, Jansen EM, Gast AT, Sher A, et al.Integrated bioinformatic and targeted deletion analyses of the SRS genesuperfamily identify SRS29C as a negative regulator of Toxoplasmavirulence. MBio. 2012;3(6):1–13.

19. Tonkin ML, Arredondo SA, Loveless BC, Serpa JJ, Makepeace KA, Sundar N,et al. Structural and biochemical characterization of Plasmodium falciparum12 (Pf12) reveals a unique interdomain organization and the potential foran antiparallel arrangement with Pf41. J Biol Chem. 2013;288(18):12805–17.

20. Myler PJ, Allison J, Agabian N, Stuart K. Antigenic variation in Africantrypanosomes by gene replacement or activation of alternate telomeres.Cell. 1984;39(1):203–11.

21. Rubio JP, Thompson JK, Cowman AF. The var genes of Plasmodiumfalciparum are located in the subtelomeric region of most chromosomes.Embo J. 1996;15(15):4069–77.

22. Freitas-Junior LH, Bottius E, Pirrit LA, Deitsch KW, Scheidig C, Guinet F, et al.Frequent ectopic recombination of virulence factor genes in telomericchromosome clusters of P. falciparum. Nature. 2000;407(6807):1018–22.

23. Marty AJ, Thompson JK, Duffy MF, Voss TS, Cowman AF, Crabb BS. Evidencethat Plasmodium falciparum chromosome end clusters are cross-linked byprotein and are the sites of both virulence gene silencing and activation.Mol Microbiol. 2006;62(1):72–83.

24. Jung C, Lee CY, Grigg ME. The SRS superfamily of Toxoplasma surfaceproteins. Int J Parasitol. 2004;34(3):285–96.

25. Carruthers VB, Tomley FM. Microneme proteins in apicomplexans. SubcellBiochem. 2008;47:33–45.

26. Rabenau KE, Sohrabi A, Tripathy A, Reitter C, Ajioka JW, Tomley FM, et al.TgM2AP participates in Toxoplasma gondii invasion of host cells and istightly associated with the adhesive protein TgMIC2. Mol Microbiol.2001;41(3):537–47.

27. Reiss M, Viebig N, Brecht S, Fourmaux MN, Soete M, Di Cristina M, et al.Identification and characterization of an escorter for two secretory adhesinsin Toxoplasma gondii. J Cell Biol. 2001;152(3):563–78.

28. Kessler H, Herm-Gotz A, Hegge S, Rauch M, Soldati-Favre D, Frischknecht F,et al. Microneme protein 8–a new essential invasion factor in Toxoplasmagondii. J Cell Sci. 2008;121(Pt 7):947–56.

29. Meissner M, Reiss M, Viebig N, Carruthers VB, Toursel C, Tomavo S, et al. Afamily of transmembrane microneme proteins of Toxoplasma gondii containEGF-like domains and function as escorters. J Cell Sci. 2002;115(Pt 3):563–74.

30. Sheiner L, Santos JM, Klages N, Parussini F, Jemmely N, Friedrich N, et al.Toxoplasma gondii transmembrane microneme proteins and their modulardesign. Mol Microbiol. 2010;77(4):912–29.

31. Opitz C, Soldati D. ‘The glideosome’: a dynamic complex powering glidingmotion and host cell invasion by Toxoplasma gondii. Mol Microbiol.2002;45(3):597–604.

32. Scholtyseck E, Mehlhorn H. Ultrastructural study of characteristic organelles(paired organelles, micronemes, micropores) of sporozoa and relatedorganisms. Z Parasitenkd. 1970;34(2):97–127.

33. Shaw MK. The same but different: the biology of Theileria sporozoite entryinto bovine cells. Int J Parasitol. 1997;27(5):457–74.

34. Hehl AB, Lekutis C, Grigg ME, Bradley PJ, Dubremetz JF, Ortega-Barria E,et al. Toxoplasma gondii homologue of plasmodium apical membraneantigen 1 is involved in invasion of host cells. Infect Immun.2000;68(12):7078–86.

35. Meissner M, Schluter D, Soldati D. Role of Toxoplasma gondii myosinA in powering parasite gliding and host cell invasion. Science.2002;298(5594):837–40.

36. Alexander DL, Mital J, Ward GE, Bradley P, Boothroyd JC. Identification ofthe moving junction complex of Toxoplasma gondii: a collaborationbetween distinct secretory organelles. PLoS Pathog. 2005;1(2):e17.

37. Beck JR, Chen AL, Kim EW, Bradley PJ. RON5 Is Critical for Organization andFunction of the Toxoplasma Moving Junction Complex. PLoS Pathog.2014;10(3):e1004025.

38. Poukchanski A, Fritz HM, Tonkin ML, Treeck M, Boulanger MJ, Boothroyd JC.Toxoplasma gondii sporozoites invade host cells using two novelparalogues of RON2 and AMA1. PLoS One. 2013;8(8):e70637.

39. Mital J, Meissner M, Soldati D, Ward GE. Conditional expression ofToxoplasma gondii apical membrane antigen-1 (TgAMA1) demonstratesthat TgAMA1 plays a critical role in host cell invasion. Mol Biol Cell.2005;16(9):4341–9.

40. Bargieri DY, Andenmatten N, Lagal V, Thiberge S, Whitelaw JA, Tardieux I,et al. Apical membrane antigen 1 mediates apicomplexan parasiteattachment but is dispensable for host cell invasion. Nat Commun.2013;4:2552.

41. Tyler JS, Boothroyd JC. The C-terminus of Toxoplasma RON2 provides thecrucial link between AMA1 and the host-associated invasion complex. PLoSPathog. 2011;7(2):e1001282.

42. Straub KW, Peng ED, Hajagos BE, Tyler JS, Bradley PJ. The moving junctionprotein RON8 facilitates firm attachment and host cell invasion inToxoplasma gondii. PLoS Pathog. 2011;7(3):e1002007.

43. Lamarque M, Besteiro S, Papoin J, Roques M, Vulliez-Le Normand B,Morlon-Guyot J, et al. The RON2-AMA1 interaction is a critical step inmoving junction-dependent invasion by apicomplexan parasites. PLoSPathog. 2011;7(2):e1001276.

44. Besteiro S, Dubremetz JF, Lebrun M. The moving junction of apicomplexanparasites: a key structure for invasion. Cell Microbiol. 2011;13(6):797–805.

45. Egarter S, Andenmatten N, Jackson AJ, Whitelaw JA, Pall G, Black JA, et al.The Toxoplasma Acto-MyoA Motor Complex Is Important but Not Essentialfor Gliding Motility and Host Cell Invasion. PLoS One. 2014;9(3):e91819.

46. Shaw MK. Cell invasion by Theileria sporozoites. Trends Parasitol.2003;19(1):2–6.

47. Hemphill A, Gottstein B, Kaufmann H. Adhesion and invasion of bovineendothelial cells by Neospora caninum. Parasitology. 1996;112(Pt 2):183–97.

48. Talevich E, Kannan N. Structural and evolutionary adaptation of rhoptrykinases and pseudokinases, a family of coccidian virulence factors. BMC EvolBiol. 2013;13:117.

49. Jensen KD, Hu K, Whitmarsh RJ, Hassan MA, Julien L, Lu D, et al. Toxoplasmagondii rhoptry 16 kinase promotes host resistance to oral infection andintestinal inflammation only in the context of the dense granule proteinGRA15. Infect Immun. 2013;81(6):2156–67.

50. Yamamoto M, Standley DM, Takashima S, Saiga H, Okuyama M, Kayama H,et al. A single polymorphic amino acid on Toxoplasma gondii kinase ROP16determines the direct and strain-specific activation of Stat3. J Exp Med.2009;206(12):2747–60.

51. Fentress SJ, Behnke MS, Dunay IR, Mashayekhi M, Rommereim LM, Fox BA,et al. Phosphorylation of immunity-related GTPases by a Toxoplasmagondii-secreted kinase promotes macrophage survival and virulence.Cell Host Microbe. 2010;8(6):484–95.

52. Steinfeldt T, Konen-Waisman S, Tong L, Pawlowski N, Lamkemeyer T,Sibley LD, et al. Phosphorylation of mouse immunity-related GTPase (IRG)resistance proteins is an evasion strategy for virulent Toxoplasma gondii.PLoS Biol. 2010;8(12):e1000576.

53. Fleckenstein MC, Reese ML, Konen-Waisman S, Boothroyd JC, Howard JC,Steinfeldt T. A Toxoplasma gondii pseudokinase inhibits host IRG resistanceproteins. PLoS Biol. 2012;10(7):e1001358.

Hehl et al. BMC Genomics (2015) 16:66 Page 16 of 16

54. Okada T, Marmansari D, Li ZM, Adilbish A, Canko S, Ueno A, et al. A noveldense granule protein, GRA22, is involved in regulating parasite egress inToxoplasma gondii. Mol Biochem Parasitol. 2013;189(1–2):5–13.

55. Rosowski EE, Lu D, Julien L, Rodda L, Gaiser RA, Jensen KD, et al.Strain-specific activation of the NF-kappaB pathway by GRA15, a novelToxoplasma gondii dense granule protein. J Exp Med. 2011;208(1):195–212.

56. Rosowski EE, Saeij JP. Toxoplasma gondii clonal strains all inhibit STAT1transcriptional activity but polymorphic effectors differentially modulateIFNgamma induced gene expression and STAT1 phosphorylation. PLoS One.2012;7(12):e51448.

57. Jensen KD, Wang Y, Wojno ED, Shastri AJ, Hu K, Cornel L, et al. Toxoplasmapolymorphic effectors determine macrophage polarization and intestinalinflammation. Cell Host Microbe. 2011;9(6):472–83.

58. Virreira Winter S, Niedelman W, Jensen KD, Rosowski EE, Julien L, Spooner E,et al. Determinants of GBP recruitment to Toxoplasma gondii vacuoles andthe parasitic factors that control it. PLoS One. 2011;6(9):e24434.

59. Gov L, Karimzadeh A, Ueno N, Lodoen MB. Human innate immunity toToxoplasma gondii is mediated by host caspase-1 and ASC and parasiteGRA15. MBio. 2013;4(4):1–11.

60. Braun L, Brenier-Pinchart MP, Yogavel M, Curt-Varesano A, Curt-Bertini RL,Hussain T, et al. A Toxoplasma dense granule protein, GRA24, modulatesthe early immune response to infection by promoting a direct andsustained host p38 MAPK activation. J Exp Med. 2013;210(10):2071–86.

61. Bougdour A, Durandau E, Brenier-Pinchart MP, Ortet P, Barakat M, Kieffer S,et al. Host cell subversion by Toxoplasma GRA16, an exported densegranule protein that targets the host cell nucleus and alters geneexpression. Cell Host Microbe. 2013;13(4):489–500.

62. Gendrin C, Bittame A, Mercier C, Cesbron-Delauw MF. Post-translationalmembrane sorting of the Toxoplasma gondii GRA6 protein into theparasite-containing vacuole is driven by its N-terminal domain. Int JParasitol. 2010;40(11):1325–34.

63. Mercier C, Adjogble KD, Daubener W, Delauw MF. Dense granules: are theykey organelles to help understand the parasitophorous vacuole of allapicomplexa parasites? Int J Parasitol. 2005;35(8):829–49.

64. Magno RC, Lemgruber L, Vommaro RC, De Souza W, Attias M. Intravacuolarnetwork may act as a mechanical support for Toxoplasma gondii inside theparasitophorous vacuole. Microsc Res Tech. 2005;67(1):45–52.

65. Coppens I, Dunn JD, Romano JD, Pypaert M, Zhang H, Boothroyd JC, et al.Toxoplasma gondii sequesters lysosomes from mammalian hosts in thevacuolar space. Cell. 2006;125(2):261–74.

66. Ferguson DJ, Jacobs D, Saman E, Dubremetz JF, Wright SE. In vivoexpression and distribution of dense granule protein 7 (GRA7) in theexoenteric (tachyzoite, bradyzoite) and enteric (coccidian) forms ofToxoplasma gondii. Parasitology. 1999;119(Pt 3):259–65.

67. Morris MT, Carruthers VB. Identification and partial characterization of asecond Kazal inhibitor in Toxoplasma gondii. Mol Biochem Parasitol.2003;128(1):119–22.

68. Morris MT, Coppin A, Tomavo S, Carruthers VB. Functional analysis ofToxoplasma gondii protease inhibitor 1. J Biol Chem. 2002;277(47):45259–66.

69. Pszenny V, Davis PH, Zhou XW, Hunter CA, Carruthers VB, Roos DS. Targeteddisruption of Toxoplasma gondii serine protease inhibitor 1 increasesbradyzoite cyst formation in vitro and parasite tissue burden in mice. InfectImmun. 2012;80(3):1156–65.

70. Balaji S, Babu MM, Iyer LM, Aravind L. Discovery of the principal specifictranscription factors of Apicomplexa and their implication for the evolutionof the AP2-integrase DNA binding domains. Nucleic Acids Res.2005;33(13):3994–4006.

71. Walker R, Gissot M, Croken MM, Huot L, Hot D, Kim K, et al. The Toxoplasmanuclear factor TgAP2XI-4 controls bradyzoite gene expression and cystformation. Mol Microbiol. 2013;87(3):641–55.

72. Radke JB, Lucas O, De Silva EK, Ma Y, Sullivan Jr WJ, Weiss LM, et al. ApiAP2transcription factor restricts development of the Toxoplasma tissue cyst.Proc Natl Acad Sci U S A. 2013;110(17):6871–6.

73. Walker R, Gissot M, Huot L, Alayi TD, Hot D, Marot G, et al. Toxoplasmatranscription factor TgAP2XI-5 regulates the expression of genes involved inparasite virulence and host invasion. J Biol Chem. 2013;288(43):31127–38.

74. Behnke MS, Wootton JC, Lehmann MM, Radke JB, Lucas O, Nawas J, et al.Coordinated progression through two subtranscriptomes underlies thetachyzoite cycle of Toxoplasma gondii. PLoS One. 2010;5(8):e12354.

75. Wang T, Zhou J, Gan X, Wang H, Ding X, Chen L, et al. Toxoplasma gondiiinduce apoptosis of neural stem cells via endoplasmic reticulum stresspathway. Parasitology. 2014;141(7):988–95.

76. Saksouk N, Bhatti MM, Kieffer S, Smith AT, Musset K, Garin J, et al.Histone-modifying complexes regulate gene expression pertinent to thedifferentiation of the protozoan parasite Toxoplasma gondii. Mol Cell Biol.2005;25(23):10301–14.

77. Lindner SE, De Silva EK, Keck JL, Llinas M. Structural determinants of DNAbinding by a P. falciparum ApiAP2 transcriptional regulator. J Mol Biol.2010;395(3):558–67.

78. Mercier C, Lefebvre-Van Hende S, Garber GE, Lecordier L, Capron A,Cesbron-Delauw MF. Common cis-acting elements critical for the expressionof several genes of Toxoplasma gondii. Mol Microbiol. 1996;21(2):421–8.

79. Cowper B, Matthews S, Tomley F. The molecular basis for the distinct hostand tissue tropisms of coccidian parasites. Mol Biochem Parasitol.2012;186(1):1–10.

80. van Dijk MR, van Schaijk BC, Khan SM, van Dooren MW, Ramesar J,Kaczanowski S, et al. Three members of the 6-cys protein family ofPlasmodium play a role in gamete fertility. PLoS Pathog. 2010;6(4):e1000853.

81. Jurankova J, Opsteegh M, Neumayerova H, Kovarcik K, Frencova A, Balaz V,et al. Quantification of Toxoplasma gondii in tissue samples ofexperimentally infected goats by magnetic capture and real-time PCR.Vet Parasitol. 2013;193(1–3):95–9.

82. Basso W, Hartnack S, Pardini L, Maksimov P, Koudela B, Venturini MC, et al.Assessment of diagnostic accuracy of a commercial ELISA for the detectionof Toxoplasma gondii infection in pigs compared with IFAT, TgSAG1-ELISAand Western blot, using a Bayesian latent class approach. Int J Parasitol.2013;43(7):565–70.

83. Rumble SM, Lacroute P, Dalca AV, Fiume M, Sidow A, Brudno M. SHRiMP:accurate mapping of short color-space reads. PLoS Comput Biol.2009;5(5):e1000386.

84. David M, Dzamba M, Lister D, Ilie L, Brudno M. SHRiMP2: sensitive yetpractical SHort Read Mapping. Bioinformatics. 2011;27(7):1011–2.

85. Anders S, McCarthy DJ, Chen Y, Okoniewski M, Smyth GK, Huber W, et al.Count-based differential expression analysis of RNA sequencing data usingR and Bioconductor. Nat Protoc. 2013;8(9):1765–86.

Submit your next manuscript to BioMed Centraland take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color figure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit